8 replies to this topic

[Mcgral18]

I'm curious about average queue times.

Alistair's recent thread thread about top 3 damaging mechs over 10 rounds inspired me to time my queues over 10 PUG matches.

I did Light mechs tonight, 10 drops. Raw times and queue % in the spoiler.

Spoiler

8PM EST

1
2:30 Light

2
2:30 14% queue Light

9:20EST

3
1:50 15% light

9:45 EST

4
4:50 20% Light

10EST

5
1:20 18% light

10:20 EST

6
0:15 16% Light

10:35 EST

7
2:50 15% Light

10:45 EST

8
3:00 15% Light

11:25 EST

9
2:30 13% Light

12 AM EST

10
1:40 20% Light

My average Light queue over 10 matches was 2 minutes and 20 seconds. Highest was 4 minutes and 50 seconds, while the lowest was a mere 15 seconds.

Match quality was...iffy. Player quality was fairly low, allowing my Cheetah to get two 1180 damage matches, and between 4 and 9 sub 200 damage players. Accurate to the average 6 sub 200.
Should probably have screen'ed all 10 matches and also noted their queue times related to player quality...perhaps for the other classes.

Group queue offered faster queues, and as expected much stiffer competition (on occasion). Those guys can actually hit light mechs...


I also thought of what was said about Release Valves on the Match Maker. How it expands every minute in the PUG LIFE.

I assume that valve is on the game instance itself, so it should be irrelevant to queue times...right?


Anyone know how the MM works in more detail?
Or perhaps you could time your own 10 PUG (or group queue) Matches with the weight class and queue% and share.

I'm curious about queue times.

[Elizander]

Mcgral18, on 06 September 2015 - 08:49 PM, said:

Group queue offered faster queues, and as expected much stiffer competition (on occasion). Those guys can actually hit light mechs...

I still fondly remember ripping off the side torso and limbs off a cheetah that tried to engage my AC/2 Jagermech. It was fun watching him run around and lose little bits and pieces of his mech until he finally ran away with about 40% of his mech left.

Seriously though, a lot of players panic when fighting lights. Being extra twitchy against them isn't good and people just don't have the patience to experiment leading shoots until they get the right distance.

Some cheetos are good enough sneaking behind to get back shots, those are the ones that get me sometimes cause I have like 3-4 back armor.

[Duke Nedo]

One feeling I have is that even if those 60 seconds are true before releasing valves, that you perhaps could get picked for another game that has somehow had "its" valves released, even though you didn't queue for 60 secs yet. I've had fast drops of any quality, can't see any trend...

Something is iffy anyways. Same with the cases when you cancel your search and it delays a bit and drops you into a game... what is that all about? It's like you have passed the point of no return...

[Elizander]

Duke Nedo, on 06 September 2015 - 10:52 PM, said:

Something is iffy anyways. Same with the cases when you cancel your search and it delays a bit and drops you into a game... what is that all about? It's like you have passed the point of no return...

Yes, there are times when I've queued for a few seconds (under 10) then decide I forgot something in my mech and cancel... then the game goes, "Oh no you don't! You signed up for a game, so you are gonna play one RIGHT NOW!"

[Wintersdark]

Mcgral18, on 06 September 2015 - 08:49 PM, said:

I also thought of what was said about Release Valves on the Match Maker. How it expands every minute in the PUG LIFE.

I assume that valve is on the game instance itself, so it should be irrelevant to queue times...right?


Anyone know how the MM works in more detail?
Or perhaps you could time your own 10 PUG (or group queue) Matches with the weight class and queue% and share.

I'm curious about queue times.

What I know about the MM is pre-PSR, so it may be incorrect now. This factor, however, should remain the same assuming release valves are even still a thing.

A key point to keep in mind regarding release valves is this:

How long you've been in the queue has no bearing on the release valves. Those release valves are timed based on how long the match has been in construction. As such, you could get into a match inside 10 seconds of searching that's been in construction for 5 minutes and thus has wide-open release valves.

[Wintersdark]

Duke Nedo, on 06 September 2015 - 10:52 PM, said:

One feeling I have is that even if those 60 seconds are true before releasing valves, that you perhaps could get picked for another game that has somehow had "its" valves released, even though you didn't queue for 60 secs yet. I've had fast drops of any quality, can't see any trend...

Something is iffy anyways. Same with the cases when you cancel your search and it delays a bit and drops you into a game... what is that all about? It's like you have passed the point of no return...

Yeah, as I said above, the release valves are timed based on match construction time, not your searching time. Remember, match construction works like this:


1) Take oldest player in the queue to seed next match.
2) Build a team at his Elo; calculate average Elo
3) Find seed player for opposing team at same rating
4) Build opposing team at that rating, calculate opposing team Elo average for comparison & end of match Elo adjustments
5) Pass teams to dedicated server for play.

Now, obviously Elo is gone and PSR is in it's place, but the process should remain effectively the same.

Steps 2 and 4, building the teams, can take a long time to find appropriate players. So, a match may well have been in construction for a couple minutes before you even clicked Play.

[Chuck Jager]

I do hit the cancel button before 2min because it can only get worse from there even if I have no precise idea of exactly what happens.

I have noticed switching weight classes and time/day makes a difference too in most areas of MM.

The rest is probably a placebo that seems to work at some level.

[Mcgral18]

Now for the Assault, a Dire Whale in this case.

Spoiler

11:22
1
3:00 Assault 27%

11:40
2
2:30 Assault 25%

11:46
3
3:45 Assault 30% 9 ******* Terribad

12 AM
4
0:40 Assault 25%

12:10
5
2:45 Assault 26%

12:26
6
2:35 Assault 21% 12 sub 200 on their side

12:38
7
2:10 Assault 20%

12:50
8
1:45 Assault 25% Cap waste of time

1:00
9
3:20 Assault 21%

1:12
10
2:45 Assault 24%

It came to 2 minutes, 30 seconds. Pretty darn close to the Light queue.
The difference between 15 and 30% queues?



Anyone else want to gather some numbers while they drop?

[kesmai]

Leveling my shadowcats I've experience longer wait times lead to lower quality teammates and the need to carry real hard. A hard task for me, cos the s-cat is not really meant to carry (my opinion).